W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2012

Re: character encoding in header fields, was: SPDY Header Frames

From: Nicolas Mailhot <nicolas.mailhot@laposte.net>
Date: Tue, 17 Jul 2012 21:05:03 +0200
Message-ID: <05ba5d67b08c2e9c87c21ed0571c145f.squirrel@arekh.dyndns.org>
To: "James M Snell" <jasnell@gmail.com>
Cc: ietf-http-wg@w3.org

> The one thing we need to determine is: how critical is the ability to
> support seamless down-level conversion from 2.0 to 1.1 within a request?
> Is it acceptable for us to say that while 2.0 can be used to transport
> 1.1 messages, the reverse is not possible.

If HTTP/2 is friendlier to mobile clients and to intermediaries, I'm quite
sure proxy implementers will want to talk to their clients in HTTP/2 only,
and CDNs like Akamai will want to translate legacy http/1 sites to modern
mobile clients (if they're not doing already that with spdy)

OTOH an HTTP/1 site will almost always expect only ascii in headers, and
for those that don't and for some reason are not ready to upgrade
protocols, it's probably acceptable to define a new way to encode i18n
strings in http/1 (prefix header values with utf8: magic prefix or
something like that). That would be a simple enough change it could be
implemented on top of existing http/1 stacks without deep changes.

(ie define a transitional http/1.2 profile close enough to current
http/1.1 it can be easily implemented, without all the warts that would
make trasformation to/from http/2 hard)

Regards,

-- 
Nicolas Mailhot
Received on Tuesday, 17 July 2012 19:05:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 17 July 2012 19:05:51 GMT